How Much Do We Really Care About Children? (Ep. 447) - Freakonomics Freakonomics
"NICKERSON: Car seats are bulky items. For the average car, the average sedan, it’s not going to be able to fit three car seats in the back. And so, the prediction would be that when I have two children that are both required to be put in car seats, it’s going to make it more difficult to have a third child...
To what degree have car seats acted as contraception? Here’s David Solomon:
SOLOMON: You can think about two numbers. What’s the change in birth rates? It’s 0.0076. So, it’s slightly less than .01 children per woman. That sounds small, but bear in mind, that’s averaging over women that have no children at all, that’s averaging over women that only have one child, or four children.
Alright, so what about the subset of people who do fall in the car-seat sweet spot?
SOLOMON: For those people, it’s about a 7.8 percent decline.
And how does that percentage decline translate to magnitude?
SOLOMON: So, if you want to talk magnitude, the easiest way to think about it is just: how many children are we talking total? In 2017, we estimate that there were about 8,000 fewer births as a result of the state of car-seat laws in the U.S...
LEVITT: The comparison between children’s car seats and children aged 2 to 6 who were wearing adult seatbelts, no matter how I slice it, I never was able to find any benefit of the car seats in terms of preventing either death or serious injury. Although it is true that in a second paper I wrote with Joe Doyle from M.I.T., we did find that child car seats are about 25 percent better at preventing the least-serious kinds of injuries in car crashes... a couple authors named Jones and Ziebarth wrote a piece in 2017 where they just tried to replicate and extend the first paper. Starting from scratch and extending the data the next 10 years in both the existing sample and in the future sample, they found exactly similar results...
We thought it might be a good idea... to run our own experiment using crash-test dummies.
LEVITT: What I thought was really incredible when we decided to go and actually put some crash-test dummies on seats and strap them into adult seatbelts and see what happened in the crashes, was the reaction of the crash test sites to doing this...
DUBNER: It was kind of a two-step amazement. No. 1 was, it was really, really hard to find a crash-test facility that would agree to do this. As it was explained to us over and over again, “All our business comes from the car-seat manufacturers. There’s no good in proving that car seats are not better than seatbelts.” And then some would say that they weren’t set up to do that. But that seemed a little bit odd because in order to do a car-seat crash, you need seatbelts. Finally, then, we found a lab that was willing to do it. The caveat was that we were not allowed to identify them by name or where they were... the guy running the lab was a scientist. And he said, “I was really intrigued by this question because it’s something I would really like to know and I think it’s good for the science.”
So, I had bought a couple of different size and brands of new car seats, and we started to hook them up and the technician who was installing the crash-test dummies in the seatbelts didn’t want to do it. There were actually two dummies — one that represented a 3-year-old child, with all these wires to measure the impact of the crash, and the other a 6-year-old dummy child. And the reason he didn’t want to do it is he was so sure that the seatbelts would fail and that the crash would break his dummy. So, then I had to sign yet another indemnification form. I was on the hook for all kinds of indemnification in case something went wrong, because this wasn’t standard testing. And then we ran the tests. And the data from the crash test basically came to the conclusion that the seatbelt had done pretty close to, if not as good a job, as the car seat. And in fact, if you had just submitted the data from the crash where the crash-test dummy was in a seatbelt alone, that would have easily passed the requirements for it to be to have been a car seat, even though it was actually just a seatbelt.
LEVITT: And I don’t know if you remember, but they were so shocked by the results that they thought there must’ve been a mistake, and they actually ran it over because they thought that something had to have gone wrong for the seatbelt to have done so well.
DUBNER: Yeah, that’s right. Now, we wrote up the results of these tests and your academic research in our Times column — and Levitt, how would you describe the reception to the news we delivered about car seats?
LEVITT: I would say somewhere between complete hatred and not even noticing. I mean, my favorite story is when the secretary of transportation at the time, Ray LaHood, came across our work. And what he writes in his blog is, “If you want to slice up the data to be provocative, have at it. As a grandfather and as secretary of an agency whose number one mission is safety, I don’t have that luxury.” I mean what the heck? His only job is to keep people safe. And when evidence comes out that maybe the policies that his department has been pushing for the last 20 years are not keeping people safe, and he has a staff of statisticians who have literally nothing else to do except to look at data to see if it’s keeping people safe, it’s amazing to me that the response isn’t, “Hey, let’s go show that these quacks are completely wrong,” or, you know, “Let’s actually try to get to the truth.” But just to be dismissive of it because it doesn’t fit with what they’ve been doing for 20 years — it was really discouraging.
DUBNER: Now, the Children’s Hospital of Philadelphia, some officials there, after we published these findings in the New York Times, wrote a letter to the editor, and it said, “Our research consistently shows that child safety seats and booster seats significantly lower the risk of serious injury compared to seatbelts alone.” They go on to say that your recommendations to rethink car-seat laws are,“irresponsible and dangerous.” So, how do you plead to that? Because you could say, “Well, maybe car seats are not as good as the authorities say.” But as you’ve said yourself, restraining of some kind for a child in a car is way better than no restraining. So, what’s the harm? Do you think you are being irresponsible and dangerous?
LEVITT: I think what’s irresponsible and dangerous is accepting mediocrity, accepting our existing solutions as if they are the best solution. What I take away from the research is we’ve put in all of these laws about child car seats and we’ve built these contraptions but then when you look at the data, they’re really not doing better, certainly not much better, than this other invention, which never had kids in mind and shouldn’t work at all...
SOLOMON: For every child’s life that’s saved by these laws, somewhere between 57 and 141 children aren’t born...
KEARNEY: So, this idea that well-meaning regulations that might have some benefit have unintended costs that could exceed the benefit is so fundamental to the way economists approach policy problems. And this is a great example of it, because it shows a place where I think nobody would have really thought of this, right? And I mean, there are so many examples of this... A glaringly obvious one that economists debate all the time, is — if we raise the minimum wage, well that looks really helpful, but are we actually going to hurt some of the people who are trying to benefit? Or here, I’ll give you another one of my current favorite policy ideas to hate, which is this idea of free college. Okay, so most people would think, how could that possibly be bad? You’re going to give free college to everyone, that’s great. Well, from the data, we know that a lot of people who go to college don’t complete. We know that the rates of completion are particularly low in nonselective schools and community colleges. You make those schools free and they have even fewer resources to devote. So, it seems an obvious unintended consequence is we’re going to diminish the quality.
It's more important to look like you're doing something useful than to actually do something useful
Of course, liberals believe humans don't respond to incentives (e.g. welfare doesn't increase single parenthood), so they don't care